如何将某些列的每组重复值放入新的分离数据帧中

import pandas as pd

# 创建示例DataFrame
data = {
    'group_id': [1, 1, 2, 2, 2, 3],
    'value': ['A', 'B', 'C', 'D', 'E', 'F']
}
df = pd.DataFrame(data)

# 查找重复值的索引
duplicates = df[df.duplicated(subset=['group_id'], keep=False)].index

# 创建一个字典来存储每个分组的数据帧
grouped_dfs = {}

for idx in duplicates:
    group_id = df.loc[idx, 'group_id']
    if group_id not in grouped_dfs:
        grouped_dfs[group_id] = df[df['group_id'] == group_id]
    else:
        grouped_dfs[group_id] = pd.concat([grouped_dfs[group_id], df.loc[[idx]]])

# 打印每个分组的数据帧
for group_id, group_df in grouped_dfs.items():
    print(f"Group ID: {group_id}")
    print(group_df)
    print("\n")

# 分块处理示例
chunk_size = 1000
chunks = pd.read_csv('large_dataset.csv', chunksize=chunk_size)

grouped_dfs = {}

for chunk in chunks:
    duplicates = chunk[chunk.duplicated(subset=['group_id'], keep=False)].index
    for idx in duplicates:
        group_id = chunk.loc[idx, 'group_id']
        if group_id not in grouped_dfs:
            grouped_dfs[group_id] = chunk[chunk['group_id'] == group_id]
        else:
            grouped_dfs[group_id] = pd.concat([grouped_dfs[group_id], chunk.loc[[idx]]])